Ray Kurzweil joins Subsense’s push for a brain interface delivered through the nose

Ray Kurzweil is joining neurotech startup Subsense as a product and vision advisor as the company develops an experimental brain interface that would deliver nanoparticles through the nose. The proposed system pairs those particles with an external wearable to read neural activity and stimulate neurons while potentially avoiding electrode implantation.

As reported in The Rundown’s September 4 newsletter, Subsense remained preclinical, had raised $27 million, and was targeting clinical pilots in 2027–29. Kurzweil’s appointment brings a prominent AI futurist to the project, but the central question remains whether the technology can work safely and reliably in people.

How Subsense’s proposed interface would work

Subsense envisions two types of nanoparticles with different jobs. One would help an external headset read neural signals; the other would stimulate neurons. Fortune’s September 3 reporting described the advisory appointment, financing, and preclinical mouse experiments.

The company’s technology page, checked September 6, describes the reading particles as plasmonic: local electric fields would change how they scatter near-infrared light. Its magnetoelectric particles would convert an applied magnetic field into local electrical effects for stimulation. These are proposed operating mechanisms; reliable delivery, reading, and stimulation together in people remain unproven.

The same roadmap targets clinical pilots in 2027–29, a clinical product in 2029–31, and a consumer brain interface from 2031 onward. Those dates are company goals. They do not establish trial schedules or a promised consumer launch. Subsense’s platform overview, also checked September 6, says its products lack FDA marketing authorization and its performance claims have not been evaluated by the agency.

Why it matters

Subsense’s potential contribution is direct neural access without the implanted electrode systems pursued by Neuralink, Synchron, and Precision Neuroscience. If that approach proves workable, it could change the procedure someone must undergo to receive a brain interface. For prospective patients and clinicians, the practical value would depend on whether avoiding electrode implantation comes with dependable function and an acceptable overall safety profile.

The comparison needs care because electrode systems differ. Neuralink’s PRIME study brochure describes surgical placement of an implant with electrode threads. Synchron’s system is delivered through the jugular vein into a blood vessel along the motor cortex. Precision’s April 2025 clearance announcement covers a cortical electrode array implanted for up to 30 days and describes a design intended for safe removal. Each presents a different procedural comparison for Subsense.

Avoiding electrode surgery would shift the safety questions toward the particles themselves. Subsense still proposes introducing material into the brain. Researchers would need to establish how consistently a dose reaches its intended destination, where else particles travel, how long they remain, and what happens with repeated administration. These are questions raised by the design. Human evidence would have to show whether the proposed delivery route offers a meaningful safety advantage.

Reliability is an equally substantial hurdle. A useful interface must repeatedly translate intended neural activity into actionable output over time. For context, Synchron reported consistent capture of motor-intent signals over 12 months in six study participants. That is a company-reported finding, but it illustrates the kind of sustained human evidence Subsense would eventually need to produce. Detecting an effect in a laboratory leaves substantial questions about everyday function.

Kurzweil’s involvement draws attention to the possibility of connecting people more directly with AI. Subsense’s own stated sequence puts medical applications before broader human–AI interfaces, giving that ambition a more immediate test: demonstrating a useful clinical function under controlled conditions. His advisory role can help shape the vision; the technology’s credibility will depend on measured human safety and performance. Prospective patients need that evidence before an easier route of administration becomes a convincing reason to choose it.